Food frequency questionnaire is a valid assessment tool of quercetin and kaempferol intake in Iranian breast cancer patients according to plasma biomarkers

Abstract

In epidemiological and clinical studies, the most common nutritional tool to assess dietary flavonol intake is the food frequency questionnaire (FFQ), which needs to contain a detailed list of plant-based foods and be previously validated. Our study aimed to assess the accuracy of dietary flavonol (quercetin, kaempferol, and isorhamnetin) intake from a food frequency questionnaire (FFQ) compared to fasting plasma flavonol concentrations, as biomarkers of exposure, in breast cancer patients. In a consecutive case series, newly diagnosed patients with breast cancer (n = 140) were recruited at Nour-Nejat Hospital, Tabriz, Iran. Flavonol intake was assessed using a validated FFQ. Plasma flavonol concentrations were measured using high-performance liquid chromatography-ultraviolet detection. The accuracy of dietary status was evaluated using a receiver operating characteristic (ROC) and area under the ROC curve (AUC). Dietary status was shown in dichotomous using ROC-cutoff point. The plasma concentrations of quercetin were moderately correlated with dietary intake of quercetin (Spearman's correlation coefficient (rs) = 0.188, P < .05; rpartial= 0.330, P < .01) and plasma concentrations of isorhamnetin (rs = 0.337, P < .001). A linear correlation between dietary levels and plasma concentrations of kaempferol was attained (rpartial = 0.240, P < .05). Using a ROC-cutoff of 61.9 nmol/L for plasma quercetin (test reference), we were able to differentiate between lower and higher consumers of quercetin with an AUCROC-based reference =0.65 (P < .01, sensitivity = 61.8%, and specificity = 60.0%). Using a plasma kaempferol concentration of 60.1 nmol/L (ROC-cutoff), it was possible to detect significant differences between higher and lower intakes of kaempferol (AUCROC-based reference = 0.64, P < .05). The correlations and diagnostic performance with plasma concentrations could present a significant accuracy rate (validity), which seems acceptable for a nutritional questionnaire (FFQ) to assess intakes intake levels of quercetin and kaempferol. An improvement in the accuracy of the flavonol exposure can provide more precise relationship with health outcomes, which may increase their clinical significance.